As a Senior Data Scientist for Ranking and Recommendations you will apply machine learning techniques such as collaborative filtering, reinforcement learning, and learning-to-rank to solve the quests of relevance prediction.
On a typical day, you will:
- build systems that delight millions of travellers,
- be a guide and mentor to your junior colleagues,
- work with the Data Analytics team to analyse exciting behavioral data and find new high-impact opportunities,
- build and tune machine learning models to improve recommendations and rankings,
- use Traveloka’s experimentation platform to track and measure your models’ success,
- effectively communicate your projects to your stakeholders and higher management, and
- be a valued voice in our effort to constantly improve our practices and frameworks.
You will own data products such as:
- Hotel recommendations using matrix-factorisation, SVD, or ALS
- Car rental search ranking using AdaRank, SVM-NDCG, or multinomial logistic regression
- Restaurant and dish recommendation using content-based filtering or pattern mining techniques such as Apriori or association rules
Working in Traveloka:
- You will work in cross-functional teams and meet great people regularly from top tier technology, consulting, product, or academic background.
- We work in an open environment where there are no boundaries or power distance.
- Everyone is encouraged to speak their mind, propose ideas, influence others, and continuously grow themselves.
- Get the exposure to multi-aspect, collaborative, intensive startup experience with our recent expansion into Southeast Asia and exploration of new products.
Required Academic Qualifications
- Masters/PhD degree from a top university in a quantitative field (Computer Science, Engineering, Physics, Mathematics or similar), or equivalent experience
- Very good theoretical understanding of relevant statistical models, their inner workings, assumptions, and limitations
- Very good understanding of evaluation metrics for relevance ranking, such as N/DCG, MAP, expected reciprocal rank, and classical MAE/MSE/RMSE-based metrics
- Academic ability to conceptualise higher-order success metrics such as serendipity and exposure diversity
Required Hands-On Experience
- 4+ years of industry experience in building recommender/ranking systems using either collaborative or content-based techniques for a consumer-facing company
- 3+ years stakeholder management skills and the ability to manage timelines and expectations
- Strong hands-on experience in the ML life-cycle for data scientists (training, testing, tuning, and performance monitoring), and a good understanding of how your friendly Data Engineering and Data Ops colleagues deploy your models to production.
- Experience with shuffling around data in cloud environments (preferably GCP: BigQuery, Pub/Sub, Dataproc) and performing the data munging required for finding new opportunities in our data.